Personalized Retail at Scale: AI-Powered Product Recommendations and Smarter Customer Service for a Growing Online Retailer
Quick glimpse
A growing online retailer needed to make shopping more relevant to each customer and maintain high support quality as case volumes rose. ABSYZ built an e-commerce website on a Salesforce Experience Site that serves personalized product recommendations and added generative AI to customer service. Cases are now handled intelligently with knowledge-based reply recommendations and automatic case summaries, improving both agent productivity and overall customer satisfaction.
Impact Delivered
- Up to 15% Lift in average order value from personalized product recommendations.
- Up to 30% Faster case resolution with knowledge-based reply recommendations.
- Up to 50% Less Time spent reviewing case history, thanks to automatic case summaries.
- Consistent Replies across every agent, drawn from the same approved knowledge base.
Tech Stack
- Salesforce Experience Cloud (Experience Site)
- Salesforce Service Cloud
- Salesforce Knowledge
- LLM / Copilot (Generative AI)
Client
- A growing direct-to-consumer retailer that sells through its own e-commerce website and supports customers through a dedicated service team.
Project Highlights
- Experience Site Storefront: Built the retailer's e-commerce website on a Salesforce Experience Site, giving customers one place to shop and get support.
- Personalized Product Recommendations: Used AI to show each customer product recommendations relevant to them, replacing a one-size-fits-all catalog.
- Knowledge-Based Reply Recommendations: Gave service agents AI-suggested replies drawn from the knowledge base, so answers are faster and more consistent.
- Automatic Case Summaries: Used an LLM-powered Copilot to summarize cases automatically, so agents understand each issue without reading the full history.
Industry
Retail / E-commerce
Challenge
In online retail, shoppers expect the store to know what they want and to provide fast answers when something goes wrong. As the retailer grew, its digital storefront and its service team both struggled to keep up with those expectations.
The core obstacles included:
- Generic Shopping Experience: Every shopper saw the same catalog, making it harder to find relevant products and holding back conversion rates.
- Missed Cross-Sell and Upsell Revenue: Without personalized suggestions, complementary products went unnoticed, and basket sizes stayed flat.
- Slow, Inconsistent Support Replies: Agents searched the knowledge base manually for each case, so response times and answer quality varied across agents.
- Time Lost on Case Context: Agents had to read lengthy case histories and email threads before responding, slowing resolution and frustrating customers.
How we helped
- Personalized E-Commerce Experience Site
We built the retailer’s online store on a Salesforce Experience Site and embedded AI-powered product recommendations throughout the shopping journey. Each shopper sees products matched to their behavior and preferences, making discovery easier and creating natural cross-sell and upsell opportunities.
- Knowledge-Based Reply Recommendations
When a customer raises a case, the AI reads the issue, finds the most relevant approved Knowledge articles, and drafts a suggested reply for the agent. Agents review, adjust if needed, and send, so every customer gets an accurate, on-brand answer in a fraction of the time.
- Automatic Case Summaries
An LLM-powered Copilot condenses each case’s history, conversations, and key details into a summary. Agents grasp the context in seconds, hand-offs between agents stay smooth, and supervisors get a clear view of every case without digging through records.
The Outcome
The retailer turned a generic online store and a manual support desk into one connected, AI-assisted customer experience.
- Shopping That Feels Personal: Every visitor gets relevant product suggestions, helping customers find what they want faster and buy more per visit.
- Faster, More Consistent Support: Agents answer cases with knowledge-backed reply suggestions, cutting response times while keeping answers accurate and on-brand.
- More Productive Service Teams: Automatic case summaries remove manual reading and note-taking, so agents handle more cases without extra headcount.